This Program Project is designed to examine the recent changes in chronic disability and functioning in the U.S. elderly populations, possible sources of those changes including the introduction of the Medicare program and, more recent, biomedical research and therapeutic innovations and the future Medicare service use and cost implications of those changes and the processes underlying them. The work builds upon a significant body of research done at CDS on the analysis and forecasting of chronic disability and health changes in the U.S. elderly population. However, that model needs to be significantly expanded in scope of adding input from economists detailed analyses of Medicare expenditures, and a more in depth analysis of specific omp0onents of health changes (e.g., dementia, stroke). To perform these analysis we have assembled a multi-disciplinary team to carry out a well integrated set of analyses. To conduct those analyses one first defines three core functions. The first (A) is longitudinally linked files but also continuous linked Medicare records which will be kept current as the project is underway (i.e., we should have Medicare data from 1999 in hand with data for 2000 shortly available thereafter; later data will be available to 2001). The third core (C) will make general health forecasts. The four projects involve (1) methodological expansion of the health model to include data from multiple sources; (2) analysis of cohort different in health and Medicare service use; (3) analyses of the natural history of Medicare expenditures; (4) analyses of the recent changes in dementia and stroke and their health cost implications These projects are all designed to take advantage of the three cores and to be integrated so that their results will help resolve the basic questions.